Struggling with M33

[Seven007]

Hello Everyone,

I am new here and this is my first post .

There is a wonderful thread on CN, which probably many already knew, were the very experienced member the elf provides his data for practicing.

https://www.cloudynights.com/topic/7496 ... rocessing/

On his homepage among others there is M33 data.
http://www.elf-of-lothlorien.de/Downloads.html

The elf achieved with PI the following result:
https://www.elf-of-lothlorien.de/Obj_Tr ... l2020.html

Elf M33 PI.JPG (111.15 KiB) Viewed 5700 times

I am unable to bring out the yellowish core, the brown dust lanes and the blue arms like him.
Mine looks like that:

Startools M33.JPG (310.63 KiB) Viewed 5700 times

Can you help me to improve my processing?

Any help is appreciated!

Best Regards
Sven

[almcl]

Hi Sven,

Non-expert disclaimer here, but this is what I get in a quick 10 minute play with the Elf's data.

I have probably overcooked the detail extraction in the HDR module and a bit more time in the color module could probably get a closer colour match, although having had a look at a couple of Hubble images, I am not sure how close to the Elf's rendition we would want to go?

M33.jpg (474.87 KiB) Viewed 5694 times

You might be interested in the many variations achieved on similar data over on Stargazers lounge. The thread here https://stargazerslounge.com/topic/3732 ... mpetition/ includes several StarTools entries and shows just how many ways there are to process the data.

[Burly]

If you post the Startools log i’am sure Ivo will be able to guide you better , makes it easier to see where maybe you could improve your work flow .

Log is in Startools folder there is a Txt file , you need to copy / paste the workflow as each time you process it creates a log .
Dave

[Seven007]

TThank you both .

I will take a look at this thread on the stargazerslonge.

I am afraid that I didn’t have the log file from the older version...

Therefore I give it a new try to get the log file information’s. This time I ended up with this. It is getting better I think, at least for me .
I still have problems with the color of the core and the dust lanes.

Startools.JPG (289.58 KiB) Viewed 5666 times

I followed a pretty simple processing path. I didn't use Decon, unfortunately it don’t know how to use it proper and my images are mostly getting worse.


StarTools 1.7.461

-----------------------------------------------------------
Image size is 6024 x 4022
---
Type of Data: Linear and was Bayered, but not whitebalanced
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [Outside RoI Influence] set to [15 %]
Parameter [RoI X1] set to [0 pixels]
Parameter [RoI Y1] set to [0 pixels]
Parameter [RoI X2] set to [6024 pixels (-0)]
Parameter [RoI Y2] set to [4022 pixels (-0)]
Parameter [Detector Gamma] set to [1.00]
Parameter [Shadow Linearity] set to [50 %]
--- Crop
Parameter [X1] set to [725 pixels]
Parameter [Y1] set to [820 pixels]
Parameter [X2] set to [5081 pixels (-943)]
Parameter [Y2] set to [3306 pixels (-716)]
Image size is 4356 x 2486
--- Bin
Parameter [Scale] set to [scale 70.71% / +1.00 bits / +0.41x SNR improvement]
Image size is 3080 x 1757
--- Wipe
Parameter [Synthetic Dark/Bias] set to [Off]
Parameter [Gradient Edge Behavior] set to [Absorb 50%]
Parameter [Synthetic Flats] set to [Off]
Parameter [Sampling Precision] set to [256 x 256 pixels]
Parameter [Dark Anomaly Filter] set to [1 pixels]
Parameter [Gradient Falloff] set to [0 %]
Parameter [Synth. Bias Edge Area] set to [100 %]
Parameter [Gradient Aggressiveness] set to [75 %]
Redoing stretch of linear data
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [2.1 pixels]
Parameter [Outside RoI Influence] set to [15 %]
Parameter [RoI X1] set to [1183 pixels]
Parameter [RoI Y1] set to [621 pixels]
Parameter [RoI X2] set to [1922 pixels (-1158)]
Parameter [RoI Y2] set to [1220 pixels (-537)]
Parameter [Detector Gamma] set to [1.00]
Parameter [Shadow Linearity] set to [50 %]
--- HDR
Parameter [Small Detail Precision] set to [Max]
Parameter [Channels] set to [Brightness Only]
Parameter [Algorithm] set to [Reveal All]
Parameter [Dark/Bright Response] set to [1.00]
Parameter [Detail Size Range] set to [1000 pixels]
Parameter [Strength] set to [1.2]
--- Wavelet Sharpen
Parameter [Structure Size] set to [Large]
Mask used (BASE64 PNG encoded)
--- SNR-aware Wavelet Sharpening
Parameter [Scale 1] set to [100 %]
Parameter [Scale 2] set to [100 %]
Parameter [Scale 3] set to [100 %]
Parameter [Scale 4] set to [100 %]
Parameter [Scale 5] set to [100 %]
Parameter [Mask Fuzz] set to [4 pixels]
Parameter [Amount] set to [400 %]
Parameter [High SNR Size Bias] set to [85 %]
Parameter [Low SNR Size Bias] set to [0 %]
Parameter [Dark/Light Enhance] set to [50% / 50%]
--- Color
Parameter [Bias Slider Mode] set to [Sliders Reduce Color Bias]
Parameter [Style] set to [Scientific (Color Constancy)]
Parameter [LRGB Method Emulation] set to [Straight CIELab Luminance Retention]
Parameter [Matrix] set to [Identity (OFF)]
Parameter [Dark Saturation] set to [2.0]
Parameter [Bright Saturation] set to [Full]
Parameter [Saturation Amount] set to [330 %]
Parameter [Blue Bias Reduce] set to [1.20]
Parameter [Green Bias Reduce] set to [1.45]
Parameter [Red Bias Reduce] set to [1.55]
Parameter [Mask Fuzz] set to [1.0 pixels]
Parameter [Cap Green] set to [0 %]
Parameter [Highlight Repair] set to [Off]
--- Unified De-Noise
Parameter [Grain Size] set to [2.0 pixels]
Parameter [Walking Noise Size] set to [1.0 pixels]
Parameter [Walking Noise Angle] set to [0]
--- Unified De-Noise
Parameter [Scale 1] set to [95 %]
Parameter [Scale 2] set to [95 %]
Parameter [Scale 3] set to [95 %]
Parameter [Scale 4] set to [95 %]
Parameter [Scale 5] set to [50 %]
Parameter [Grain Equalization] set to [0 %]
Parameter [Scale Correlation] set to [50 %]
Parameter [Color Detail Loss] set to [50 %]
Parameter [Brightness Detail Loss] set to [50 %]
Parameter [Grain Dispersion] set to [2.0 pixels]

[Russ.Carpenter]

I'm not sure the world needs another pic of M33, but couldn't resist taking a run at this data. The problem with M33 is that it's your basic blob. The question is: how much local contrast should I force upon it?

Russ

[Seven007]

Hi Russ,

I cannot tell you if the world need another pic of M33. But if it is needed it will look like your version.

Thank you, very nice Image. Please show me your workflow.

Best Regards
Sven

[admin]

Hi Sven,

Processing M33 is a great example of why letting the data guide you is a superior way to processing, rather than forcing any preconceptions onto your dataset with blunt force instruments to effectuate the detail you assume is there.

As Russ and others remarked, M33 is somewhat featureless/boring in terms of grand structures. It has no clear bar structure and it lacks a typical core/bulge. Its (messy) spiral structure with many spurs, emanates directly from the core and not from some inner region/ring, There a no "neat" dust lanes due to the many spurs - regions of dust are mostly "broken up".

Of course at the end of the day, as others also remarked, you can push local contrast as much as you see fit, to flatten the disc (use the Locality parameter in the Contrast module), and use the HDR module to accentuate the spurs and/or reveal details in the core.

[Seven007]

Hi Ivo,
Thank you for the information of M33 and processing .

I have to apologize for my way to put this in here which leads to wrong intentions of me what I like to do.

My main goal was to learn and understand Startools, M33 was “just” my data I choose from the thread of CN.

The other members on CN and me asked how to process this to come to a given result. One answer was it could be explained for PI but not for Startools the PI user of course don’t know the modules to use in Startools.

Very often a general answer is, if you like to process you have to use PI. I saw a yellow core and liked to know how to get there as a lesson in processing. if it is possible in PI then why not also in Startools? I like Startools so I tried to figure out and this leads to my thread here.

It was never my intention to force something what might not be there. More to understand what the Modules are used for and what happens if I change the settings. I am a beginner in Startools and like to understand and not blindly use it.

I really like the idea of its signal tracking and data use as well how you argue in the different forums about it. I am from a scientific background and it reminds me sometimes a little bit of Quality by Design, Design of Experiments and Six sigma.

You are right about data which guides, in my case I am learning where it wants to guide me to. So sorry again for putting this into a wrong light I just want to learn processing .

Best Regards and thanks

[Russ.Carpenter]

Hi Sven, again,

Ivo gave you a really useful answer. Locality and HDR are the main tools here. They are spectacular features of StarTools, but sometimes it's hard to know where to stop! I'll mention that I had to fiddle with the color module a lot. Sometimes, the first pass with this module gives a credible result, but sometimes it doesn't. That's what the RGB sliders are for. Occasionally I forget to check for residual green with Max RGB. That's a mistake. Green usually needs to be nuked.

Russ

[admin]

Hi Ivo,
Thank you for the information of M33 and processing .

I have to apologize for my way to put this in here which leads to wrong intentions of me what I like to do.

My main goal was to learn and understand Startools, M33 was “just” my data I choose from the thread of CN.

The other members on CN and me asked how to process this to come to a given result. One answer was it could be explained for PI but not for Startools the PI user of course don’t know the modules to use in Startools.

Very often a general answer is, if you like to process you have to use PI. I saw a yellow core and liked to know how to get there as a lesson in processing. if it is possible in PI then why not also in Startools? I like Startools so I tried to figure out and this leads to my thread here.

It was never my intention to force something what might not be there. More to understand what the Modules are used for and what happens if I change the settings. I am a beginner in Startools and like to understand and not blindly use it.

Please never apologize for asking any and all questions!

The dataset in question is a bit... odd and not pre-processed according to recommended/best practices. Somewhere in the long thread I asked about it and - if I remember correctly - the Elf who so graciously made it available for all of us, mentioned, amongst other things, he is intentionally coloring the flats (which if calibrated per-color channel will greatly influence the signal's prevalence and quality).

Have a look at the M33 IKI Observatory datasets (links & tutorials section) for some very high quality data, which will help with a baseline for expectations for this object.

Neither his PI nor any ST rendition I have produced myself make me very comfortable when it comes to the coloring. Something is off and there a quite a few pre-processing variables we don't (didn't) have control over.

Indeed, yellow cores are expected features of many galaxies. One of the main reasons for this can be found, for example, on the Wikipedia page for M33. Under star formation it is stated;

In the central 4′ region of this galaxy, atomic gas is being efficiently converted to molecular gas, resulting in a strong spectral emission of CO. This effect occurs as giant molecular clouds condense out of the surrounding interstellar medium. A similar process is taking place outside the central 4′, but at a less efficient pace. About 10% of the gas content in this galaxy is in the molecular form.[36][37]

Star formation is taking place at a rate that is strongly correlated with the local gas density, and the rate per unit area is higher than in the neighboring Andromeda Galaxy. (The rate of star formation is about 3.4 Gyr−1 pc−2 in the Triangulum Galaxy, compared to 0.74 in Andromeda.[44]) The total integrated rate of star formation in the Triangulum Galaxy is about 0.45 ± 0.1 solar masses per year. It is uncertain whether this net rate is currently decreasing or remaining constant.[36][37]

Based on analysis of the chemical composition of this galaxy, it appears to be divided into two distinct components with differing histories. The inner disk within a radius of 30×103 ly (9 kpc) has a typical composition gradient that decreases linearly from the core. Beyond this radius, out to about 82×103 ly (25 kpc), the gradient is much flatter. This suggests a different star formation history between the inner disk and the outer disk and halo, and may be explained by a scenario of "inside-out" galaxy formation.[38] This occurs when gas is accumulated at large radii later in a galaxy's life space, while the gas at the core becomes exhausted. The result is a decrease in the average age of stars with increasing radius from the galaxy core.[45]

E.g. in the core of M33 (and galaxies like it), older stars remain, while in the outer regions younger stars are still present.
The massive blue stars (O and B-class) live fast and die young (usually in a spectacular fashion). So in older areas, they are no longer present, skewing the coloring to comparatively long-lived red and yellow stars.

I really like the idea of its signal tracking and data use as well how you argue in the different forums about it. I am from a scientific background and it reminds me sometimes a little bit of Quality by Design, Design of Experiments and Six sigma.

Thank you so much! Quality by Design / DFSS is indeed very closely related to ST's approach. IMHO excellence comes from thoroughly understanding and measuring the entire process wherever objectively possible. Astronomical signal processing is far more than the sum of its parts, and this can be used to the enhance the process significantly to the user's advantage (features that create joy/satisfaction and predictability in the way they work).

You are right about data which guides, in my case I am learning where it wants to guide me to.

Good data is extremely helpful in that regard. Particularly having multiple datasets on the same object, and having a workflow that is somewhat replicable with defaults you can trust.

I can still remember a long time ago when I was processing M81 and noticed this streak across the core. It looked like a dust lane but was criss-crossing the core. I thought it must have been a flats issue. I kept it as-it-was, trusting the data regardless. I then processed another M81 dataset. It had the same thing. It's one of those "huh - that's interesting" moments. I then started researching it and it is a real (but peculiar) feature. It has some hints of dust lanes that appear to cross the disc, as a result of (possibly) interaction with M82 or NGC 3077.

When you repeat an "experiment" (processing) with data acquired by different people with comparable instruments, and you get the same results, that is science!